Clerq's ten-minute patentability report targets the junior-associate line item

Toronto patent research startup NLPatent renamed itself Clerq and shipped its first agentic workflows, which it says run a full patentability analysis in about 10 minutes with citations attached to every conclusion ^[1][2]. The company's own framing of the before state is the more interesting number: that work normally takes days or weeks, and firms usually push it down to a junior associate or out to a search firm ^[3].

That is a line item, not a task. If a day of billable work is eight hours, a 10-minute run is roughly one forty-eighth of the smallest version of the old job ^[4]. The unit being automated is not "research" in the abstract; it is the delegated hour, whether it sat inside the firm at associate rates or outside it on a search vendor invoice. Chief executive and co-founder Stephanie Curcio put it as a category shift: "This isn't a smarter version of the same software-as-a-service model, it's service as software," and said work that used to be delegated down or sent out now happens in minutes, under the direction of the professional accountable for it ^[5].

The accountability seam is where the product lives or dies. Two workflows are live: a rapid triage assessment for screening high volumes of invention disclosures, and a full patentability report with feature-by-feature reasoning and cited references ^[6]. Clerq says an attorney reviews the reasoning before signing off ^[7]. Feature-by-feature reasoning and cited references are what make that review possible at all; a conclusion without a traceable path is not reviewable, it is only readable. The name itself encodes the positioning, referring to a law clerk who does substantive work under an attorney's direction ^[8]. Whether the seam holds depends on how long attorney review actually takes when the output is wrong in a way that looks right.

Two signals suggest the company knows which buyer it is displacing. It hired Michael Chernoff as director of IP strategy; he began as a U.S. Patent and Trademark Office examiner and went on to run a 30-person patent search practice at Murgitroyd, with a stint as chief patent counsel at a Fortune 500 company and chief intellectual property officer at an investment bank in between ^[9]. And it announced cross-referral integrations that put its research engine inside RPX Empower, which RPX Corp. sells to companies managing large litigation exposure, and gives Park IP, a Welocalize company, the same research layer for patent translation and foreign filing work ^[10]. RPX Chief IP and SaaS Officer Steve Chiang said patent risk assessment "hinges on the quality of the patent research supporting each determination" ^[11].

Context worth keeping: law firms, corporate IP departments and universities already use the company's large language models for search and monitoring, and those models predate ChatGPT ^[12]. Clerq runs its models in a private cloud and says it does not train on customer data ^[13]. It was founded in 2021 by Curcio and James Stonehill and raised $3 million in November in a round led by Draper Associates and Mighty Capital, money Curcio told BetaKit would go toward exactly these agentic workflows ^[14][15].

Watch the roadmap items, invalidity and freedom-to-operate ^[16], because those carry litigation consequence rather than filing-decision consequence, and the review burden scales with the stakes. Watch also whether buyers price this as a software seat or as displaced search spend. Service as software only means something if someone's outsourced-search budget line shrinks.